NVIDIA-Merlin / NVIDIA-Merlin/Transformers4Rec
Investigate approaches to distribute huge embedding tables in PyTorch
Open
Nobody has claimed this yet.
area/pytorch
scalability
- Dominant language
- Python
- Stars
- 1.3k
- Forks
- 165
- Avg merge
- 1m
- Merged PRs (30d)
- 2
Description
- Investigate the alternatives to have distributed embedding tables with native PyT
- Check (if possible) how we can leverage HugeCTR engine for PyT / TF for distributed embeddings
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by investigating native PyTorch approaches for distributing embedding tables, then check whether HugeCTR can be leveraged for distributed embeddings in PyTorch or TensorFlow. Done means documenting the viable alternatives and the feasibility of HugeCTR integration.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python, pytorch
- Domain
- distributed-systems, machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 25/100